The unfriendly skies

In April of 2010 the federal government, faced with numerous horror stories of airplanes languishing on airport tarmacs for many hours, imposed a new regulation whereby airlines would be forced to let passengers off of planes sitting on the tarmac for three hours or would face fines of up to $27,500 per passenger. It's a typical example of simplistic bureaucrat/politician thinking: identify a problem (preferably one that is high-profile and will generate positive coverage), pass a law banning it, then consider the problem solved.

The more sophisticated approach would be to identify the dynamics at play in the country's air transport system that resulted in such delays and how they might be resolved. After all, it was not as if airlines abused passengers for the fun of it or had any self-interest in letting expensive assets sit on the ground rather than be put to productive use. Certainly there must have been rules and incentives in place that produced this poor outcome for both airlines and consumers that could have been examined and changed. Careful analysis, however, is difficult whereas simply imposing new regulations is easy.  

The sophisticated approach would also take into consideration not only the obvious benefits of a certain course of action, but its costs as well. For example, in assessing the impact of minimum wage increases one must consider not only the higher wages that result but also secondary effects such as increased unemployment, businesses reducing investment elsewhere, etc. In the case of the new airline rule, the obvious cost was that airlines were now given a massive incentive to cancel flights in danger of experiencing three hour delays rather than run the risk of incurring a fine that could easily reach into the millions of dollars. That passengers are better served by canceled flights rather than delays is not obvious.

Now that we are roughly at the two year anniversary of this rule, it's worth examining how it has played out. Last summer the Department of Transportation issued a press release hailing the regulation's success:

During the first 12 months after a new rule limiting airline tarmac delays went into effect, lengthy delays experienced by passengers aboard aircraft largely disappeared and only a minimal number of flights were canceled to avoid delays on the tarmac, the U.S. Department of Transportation (DOT) announced today. 

“On the one-year anniversary of the tarmac delay rule, it’s clear that we’ve accomplished our goal of virtually eliminating the number of aircraft leaving travelers stranded without access to food, water, or working lavatories for hours on end,” U.S. Transportation Secretary Ray LaHood said.  “This is a giant step forward for the rights of air travelers.”

The press release cites the following numbers to support its claim of eliminating tarmac delays with only a minimal impact on flight cancelations:

During April, the carriers canceled 2.0 percent of their scheduled domestic flights, compared to 0.7 percent in April 2010 and 1.3 percent in March 2011.

The decision to use percentage points rather than absolute figures smells like a case of lies, damned lies and statistics. According to this Bureau of Transportation press release, in January 2011 there were 680,300 domestic flights for that one month. If we conservatively assume that April of that year saw 600,000 flights, and 2 percent of them were canceled, that's 12,000 canceled flights.

In comparison, one year earlier the number of canceled flights for the same month was 0.7 percent, or 4,200 flights. Given that it's likely the number of canceled flights increased by roughly 8,000, the DOT's decision to opt for percentages over absolute numbers is understandable (other observers have also noted the department's cherry picking of data).

This, of course, is using back-of-the-envelope calculations and is speculative. However, the Government Accountability Office released a more rigorous analysis last September which concluded:

The rule appears to be associated with an increased number of cancellations for thousands of additional passengers—far more than DOT initially predicted—including some who might not have experienced a tarmac delay.

The report also found that US airlines were 24 percent more likely to more than three times as likely to cancel a flight in 2010 after the three-hour tarmac delay rule went into effect, compared with the previous year.

Lastly, let's note that libertarians -- that small faction of the electorate widely derided by both Democrats and Republicans as the crazy aunts in the attic -- predicted this very phenomenon of increased cancellations (in fairness, anyone who gave the issue even a cursory examination could see this coming). There are, however, no apparent signs of introspection or analysis from the progressive left over this rule change -- no doubt they have already moved on and are devoting their energies to dreaming up new rules and regulations to further burden the economy with.